The utility of primary human hepatocytes in the evaluation of drug-drug interactions is being investigated in our laboratories. Our initial approach was to investigate whether drug-drug interactions observed in humans in vivo could be reproduced in vitro using human hepatocytes. Two model drugs were studied: terfenadine and rifampin, representing compounds subjected to drug-drug interactions via inhibitory and induction mechanisms, respectively. Terfenadine was found to be metabolized by human hepatocytes to C-oxidation and N-dealkylation products as observed in humans in vivo. Metabolism by human hepatocytes was found to be inhibited by drugs which are known to be inhibitory in vivo. K-i values for the various inhibitors were derived from the in vitro metabolism data, resulting in the following ranking of inhibitory potency: For the inhibition of C-oxidation, ketoconazole > itraconazole > cyclosporin similar to troleandomycin > erythromycin > naringenin. For the inhibition of N-dealkylation, itraconazole greater than or equal to ketoconazole > cyclosporin greater than or equal to naringenin greater than or equal to erythromycin greater than or equal to troleandomycin. Rifampin induction of CYP3A, a known effect of rifampin in vivo, was also reproduced in primary human hepatocytes. Induction of CYP3A4, measured as testosterone 6 beta-hydroxylation, was found to be dose-dependent, treatment duration-dependent, and reversible. The induction effect of rifampin was observed in hepatocytes isolated from all 7 human donors studied, with ages ranging from 1.7 to 78 years. To demonstrate that the rifampin-induction of testosterone 6 beta-hydroxylation could be generalized to other CYP3A4 substrates, we evaluated the metabolism of another known substrate of CYP3A4, lidocaine. Dose-dependent induction of lidocaine metabolism by rifampin is observed. Our results suggest that primary human hepatocytes may be a useful experimental system for preclinical evaluation of drug-drug interaction potential during drug development, and as a tool to evaluate the mechanism of clinically observed drug-drug interactions.